Files
stalwart-migrator/internal/stalwartapi/management.go
T
jcoffey-dev 4568f9abbf Capture the pre-migration snapshot from 0.15.x, and stop claiming counts match when none were compared
Found by running preflight against a real Stalwart 0.15.5 in a VM. Two
defects, the second worse than the first.

1. AccountSnapshot could not read the version this tool migrates FROM.
   0.15.5 advertises no urn:stalwart:jmap capability and POST /api returns
   404 - the JMAP management API and x:Account are 0.16 features. 0.15.x
   exposes a REST API at GET /api/principal instead. So preflight's
   account-snapshot check warned and moved on, and every run against a real
   source instance had no "before" data at all.

   AccountSnapshot now dispatches on the capability the session document
   advertises - a positive signal, not an inference from a failed call -
   and internal/stalwartapi/principal.go implements the 0.15.x REST path,
   including its 1-based page/limit pagination so an install larger than
   one page isn't silently truncated.

2. With no "before" counts, the content-integrity comparison iterated an
   empty map, checked nothing, and reported "all message counts match".
   That is the strongest claim this tool makes - ARCHITECTURE 4.7 calls it
   the actual no-data-loss guarantee - made vacuously, and it would have
   passed on a migration that lost every message.

   The comparison now derives its account set from whatever the source
   could report, verifies every account and domain survived either way, and
   carries MessageCountsCompared so the report says plainly "MESSAGE COUNTS
   NOT COMPARED ... no-data-loss is NOT verified here" rather than implying
   otherwise.

What can and cannot be checked across the 0.15/0.16 boundary, now that a
real server has answered: 0.15.x has no per-mailbox message count at any
endpoint, and the impersonation login 0.16 offers returns 401 there, so
before/after message counts are impossible for the boundary migration this
tool exists for. Both versions do report per-account used quota (usedQuota
in 0.15's REST list, usedDiskQuota on 0.16's x:Account), so that is
captured on both sides. It is recorded and reported, not asserted on:
4.5 notes the 0.16 migration resets quotas to zero pending recalculation,
so comparing those bytes across the boundary would be a false alarm
generator.

Test servers across preflight, validate and stalwartapi now advertise
urn:stalwart:jmap, since they stand in for 0.16 instances and that
capability is what says so.

Verified end to end against the smoke VM: all nine preflight checks pass,
and the checkpoint records 2 accounts, 1 domain and per-account used quota
where it previously recorded nothing.
2026-08-23 18:51:19 -07:00

241 lines
8.3 KiB
Go

// SPDX-FileCopyrightText: 2026 LINUXexpert-org
// SPDX-License-Identifier: GPL-3.0-or-later
package stalwartapi
import (
"bytes"
"context"
"encoding/json"
"fmt"
"io"
"net/http"
"sort"
"strings"
)
// managementCapabilities are the JMAP capability URNs Stalwart requires for
// its management object calls (x:Account/*): standard JMAP core plus its
// own urn:stalwart:jmap extension. Confirmed against
// docs/ref/object/account.md and crates/jmap-proto/src/request/capability.rs
// in stalwartlabs/stalwart.
var managementCapabilities = []string{"urn:ietf:params:jmap:core", "urn:stalwart:jmap"}
type jmapRequest struct {
Using []string `json:"using"`
MethodCalls []any `json:"methodCalls"`
}
type jmapRawResponse struct {
MethodResponses []json.RawMessage `json:"methodResponses"`
}
// methodResponse is one [name, args, callId] triple from a JMAP response,
// per RFC 8620 §3.2 - Stalwart's management API follows the same envelope
// shape as its regular JMAP methods.
type methodResponse struct {
Name string
Args json.RawMessage
CallID string
}
// call POSTs one JMAP-style request to the management API (Stalwart's /api
// endpoint - see docs/ref/object/account.md, which is distinct from /jmap)
// using this Client's own credentials, and returns its parsed method
// responses in order.
func (c *Client) call(ctx context.Context, using []string, methodCalls []any) ([]methodResponse, error) {
return c.callAs(ctx, c.Username, c.Password, strings.TrimRight(c.BaseURL, "/")+"/api", using, methodCalls)
}
// callAs is call's underlying primitive: it accepts an explicit
// username/password/URL rather than always using this Client's own
// credentials and the management endpoint. MailboxSnapshot uses this to
// call standard JMAP methods (not Stalwart's x: management objects) against
// the URL a JMAP session document says to use, authenticated as an
// impersonated identity rather than this Client's own.
func (c *Client) callAs(ctx context.Context, username, password, url string, using []string, methodCalls []any) ([]methodResponse, error) {
reqBody, err := json.Marshal(jmapRequest{Using: using, MethodCalls: methodCalls})
if err != nil {
return nil, fmt.Errorf("stalwartapi: encode request: %w", err)
}
req, err := http.NewRequestWithContext(ctx, http.MethodPost, url, bytes.NewReader(reqBody))
if err != nil {
return nil, err
}
req.Header.Set("Content-Type", "application/json")
req.SetBasicAuth(username, password)
resp, err := c.httpClient().Do(req)
if err != nil {
return nil, fmt.Errorf("stalwartapi: call %s: %w", url, err)
}
defer resp.Body.Close()
if resp.StatusCode != http.StatusOK {
body, _ := io.ReadAll(io.LimitReader(resp.Body, 4096))
return nil, fmt.Errorf("stalwartapi: %s returned %s: %s", url, resp.Status, strings.TrimSpace(string(body)))
}
var raw jmapRawResponse
if err := json.NewDecoder(resp.Body).Decode(&raw); err != nil {
return nil, fmt.Errorf("stalwartapi: decode response from %s: %w", url, err)
}
responses := make([]methodResponse, 0, len(raw.MethodResponses))
for _, r := range raw.MethodResponses {
var triple [3]json.RawMessage
if err := json.Unmarshal(r, &triple); err != nil {
return nil, fmt.Errorf("stalwartapi: parse method response envelope: %w", err)
}
var name, callID string
if err := json.Unmarshal(triple[0], &name); err != nil {
return nil, fmt.Errorf("stalwartapi: parse method response name: %w", err)
}
if err := json.Unmarshal(triple[2], &callID); err != nil {
return nil, fmt.Errorf("stalwartapi: parse method response call id: %w", err)
}
responses = append(responses, methodResponse{Name: name, Args: triple[1], CallID: callID})
}
return responses, nil
}
// account is the subset of x:Account/get's response fields this tool needs,
// confirmed against Stalwart's own docs/ref/object/account.md (whose
// stalwart-cli example is `query Account --fields id,name,domainId,usedDiskQuota`).
type account struct {
ID string `json:"id"`
Name string `json:"name"`
DomainID string `json:"domainId"`
// UsedDiskQuota is 0.16's name for what 0.15's REST API calls
// usedQuota - the same per-account byte count, which is what makes a
// cross-boundary content comparison possible at all.
UsedDiskQuota int64 `json:"usedDiskQuota"`
}
// AccountSnapshot enumerates every account on the instance via Stalwart's
// management API - x:Account/query to list ids, then x:Account/get to fetch
// their name/domainId - and returns the account count, set of domains in
// use, and (via MailboxSnapshot, per account) every mailbox's message
// count. This is what preflight's snapshot (ARCHITECTURE.md §4.1) and
// validate's directory-integrity and content-integrity checks (§4.7)
// compare before and after migration - the latter being the actual
// no-data-loss guarantee.
//
// A per-account mailbox-count failure (most likely: this Client's Username
// lacks the `impersonate` permission MailboxSnapshot depends on) does not
// fail the whole snapshot - the account/domain enumeration above is already
// useful on its own, and one account's failure shouldn't hide a working
// result for every other account. Instead it's recorded in
// Snapshot.MailboxErrors, keyed by account email, so callers can report
// exactly what's missing rather than silently treating an unreachable
// account's mailboxes as having zero messages.
func (c *Client) AccountSnapshot(ctx context.Context) (*Snapshot, error) {
isJMAP, err := c.hasJMAPManagement(ctx)
if err != nil {
return nil, fmt.Errorf("stalwartapi: discover which management API this instance speaks: %w", err)
}
if !isJMAP {
// No urn:stalwart:jmap: this is a 0.15.x instance, whose
// management API is REST. Confirmed against a live 0.15.5 server -
// see principal.go.
return c.principalSnapshotREST(ctx)
}
return c.accountSnapshotJMAP(ctx)
}
// accountSnapshotJMAP is the 0.16+ path, via Stalwart's JMAP management
// objects.
func (c *Client) accountSnapshotJMAP(ctx context.Context) (*Snapshot, error) {
ids, err := c.AccountIDs(ctx)
if err != nil {
return nil, err
}
if len(ids) == 0 {
return &Snapshot{}, nil
}
getResp, err := c.call(ctx, managementCapabilities, []any{
[]any{"x:Account/get", map[string]any{"ids": ids, "properties": []string{"id", "name", "domainId", "usedDiskQuota"}}, "g"},
})
if err != nil {
return nil, fmt.Errorf("stalwartapi: Account/get: %w", err)
}
accounts, err := accountGetList(getResp)
if err != nil {
return nil, err
}
mailboxCounts := map[string][]MailboxCount{}
mailboxErrors := map[string]string{}
for _, a := range accounts {
if a.Name == "" {
continue // no login/email to impersonate against
}
counts, err := c.MailboxSnapshot(ctx, a.Name)
if err != nil {
mailboxErrors[a.Name] = err.Error()
continue
}
mailboxCounts[a.Name] = counts
}
domainSet := map[string]bool{}
for _, a := range accounts {
if a.DomainID != "" {
domainSet[a.DomainID] = true
}
}
domains := make([]string, 0, len(domainSet))
for d := range domainSet {
domains = append(domains, d)
}
sort.Strings(domains)
usedQuota := make(map[string]int64, len(accounts))
for _, a := range accounts {
if a.Name != "" {
usedQuota[a.Name] = a.UsedDiskQuota
}
}
return &Snapshot{
AccountCount: len(accounts),
Domains: domains,
MailboxCounts: mailboxCounts,
MailboxErrors: mailboxErrors,
UsedQuota: usedQuota,
}, nil
}
func accountQueryIDs(responses []methodResponse) ([]string, error) {
if len(responses) == 0 {
return nil, fmt.Errorf("stalwartapi: Account/query returned no method responses")
}
r := responses[0]
if r.Name == "error" {
return nil, fmt.Errorf("stalwartapi: Account/query error: %s", r.Args)
}
var result struct {
IDs []string `json:"ids"`
}
if err := json.Unmarshal(r.Args, &result); err != nil {
return nil, fmt.Errorf("stalwartapi: parse Account/query response: %w", err)
}
return result.IDs, nil
}
func accountGetList(responses []methodResponse) ([]account, error) {
if len(responses) == 0 {
return nil, fmt.Errorf("stalwartapi: Account/get returned no method responses")
}
r := responses[0]
if r.Name == "error" {
return nil, fmt.Errorf("stalwartapi: Account/get error: %s", r.Args)
}
var result struct {
List []account `json:"list"`
}
if err := json.Unmarshal(r.Args, &result); err != nil {
return nil, fmt.Errorf("stalwartapi: parse Account/get response: %w", err)
}
return result.List, nil
}